门式刚架的振动损伤识别方法研究  

Reseach on portal steel frame's damage indentification methods by vibrations

作　　者：贾宏玉[1,2] 郝宏伟[3] 岳鹏飞[1] 

机构地区：[1]内蒙古科技大学建筑与土木工程学院,内蒙古包头014010 [2]南京航空航天大学宇航学院,江苏南京210016 [3]西安建筑科技大学土木工程学院,陕西西安710055

出　　处：《内蒙古科技大学学报》2011年第3期278-281,共4页Journal of Inner Mongolia University of Science and Technology

基　　金：内蒙古科技大学创新基金资助项目(2009NC040)

摘　　要：应用ANSYS建立槽型截面门式刚架无损伤和有损伤模型,求解无损伤和有损伤情况下的门式刚架固有频率.应用Matlab软件作出门式刚架无损伤时和有损伤时频率变化的对比曲线图.采用BP神经网络对损伤识别和定位进行模拟研究,使神经网络能够通过固有频率识别出相应的损伤位置和程度.仿真结果显示,BP神经网络用于结构损伤识别有很大优势.The no-damage model and various damage models of channel section portal steel frame were established using the software ANSYS, and their natrual frequencies were solved. The changes of natural frequencies curves of no-damage model in contrast with various damage models were ploted with the software Matlab. The simulation research was carried out on damage identification and location by BP neural network, and it is concluded that the neural network can identify the position and the degree of damage by natural frequencies. The simulated results proved the advantage of using BP neural network for identifying the damage of the structure.

关 键 词：门式刚架 损伤识别 固有频率变化 BP神经网络 

分 类 号：TU328[建筑科学—结构工程]